|Publication number||US4590554 A|
|Application number||US 06/443,937|
|Publication date||May 20, 1986|
|Filing date||Nov 23, 1982|
|Priority date||Nov 23, 1982|
|Also published as||EP0143125A2|
|Publication number||06443937, 443937, US 4590554 A, US 4590554A, US-A-4590554, US4590554 A, US4590554A|
|Inventors||Sam D. Glazer, James Baumbach, Anita Borg, Emanuel Wittels|
|Original Assignee||Parallel Computers Systems, Inc.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (8), Referenced by (148), Classifications (18), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This invention pertains to multiprocessor fault tolerant computer systems.
All fault tolerant computer systems require the duplication of both hardware and software resources. These systems differ in the way that the duplicate resources are used to provide fault tolerance and in the amount of additional computing power gained from the additional hardware. Such systems are similar in that they require that every primary process or task (executing program) has a backup process or task in a backup means capable of continuing execution if the primary task fails.
One such approach (e.g., STRATUS systems) requires that a process or task and its backup execute simultaneously on two CPUs (central processing unit) so that if one fails the other continues without interruption. One of the CPUs is generally called a "primary" and the second is generally called a "secondary" or "backup". The primary and backup are essentially indistinguishable. While such a system provides for fault tolerance, it should be noted that the duplicate hardware provides no increased computational capability.
A second approach involves keeping an inactive (nonexecuting) backup process which, in the case of failure of the primary CPU can be brought up on another backup or secondary CPU. In such case, the backup task must be kept in such a way that at any time it can take over from its primary task. In other words, the state of the backup task as represented by the universe or collection of values in its data space, must be either equal or identical to that of the primary task or be capable of becoming identical. In such systems, as long as failure does not occur, the duplicate hardware in the backup CPU can be used to run additional primary processes.
Such a system is shown in U.S. Pat. No. 4,228,496, later implemented by the assignee in the Tandem series of computer systems. In such a system, whenever the primary task receives a message, the primary task will now perform a computation based on the newly received message and produce new data which is stored in its date space resulting in a change thereof. When this data space changes, a checkpointing operation may be performed wherein the data space associated with the primary task is made available as the data space of the secondary or backup task. Thus, when a failure occurs in the primary, the backup task will continue to operate with the current data and information that was provided by the checkpointing of the primary task. Although in this type of a system the backup task is inactive so that the extra hardware can be used for other purposes, the frequent copying of the primary's data space into the secondary's data space often used a large portion of the added computing power.
It is a general object of the invention to provide for fault tolerance while at the same time increasing the efficiency of the overall operation during normal processing by reducing the frequency of the large transfers of data and information required during checkpointing.
To accomplish this object, the invention contemplates that instead of keeping the backup or secondary task exactly up to date, the backup is kept nearly up to date but is provided with all information necessary to bring itself up to the state of the primary task should there by a failure of the primary task. The inventive concept is based on the notion that if two tasks start out in identical states and are given identical input information, they will perform identically.
In particular, all inputs to a process running on a system according to the invention are provided via messages. Therefore, all messages sent to the primary task must be made available to the secondary or backup task so that upon failure of the primary task the secondary task catches up by recomputing based on the messages. In essence, then, this is accomplished by allowing every backup task to "listen in on" its primary's message.
According to further features of the invention which avoid the complete recomputation by the backup task upon failure, a primary task and its backup task are periodically "synchronized" by a copy of the primary task's data space. While this synchronizing includes the checkpointing described above, it occurs less frequently. In the intervening periods, when the backup task is not identical to the primary, all messages to the primary task are kept available for the backup task. If the primary task then fails, the backup executes from the point of last synchronization using the saved input.
It should be apparent that such a routine makes normal computation much more efficient at the expense of a small amount of additional computation should a failure actually occur.
In addition, the method according to the invention requires fewer message transfers than the heretofore known checkpointing methods. There immediately also follows the fact that less computation time is required and less memory space needed as compared to conventional checkpointing.
More specifically, the invention contemplates a parallel computer system having at least a first and a second primary task performing means and a first and a second secondary (backup) task performing means. The task performing means are interconnected by a message bus means. In such a system, there is contemplated the method of sending messages among the task performing means. In particular, the invention contemplates sending messages from the first primary task performing means to the second primary task performing means which operates on the messages in accordance with the task associated with such means. The second primary task performing means operates on the received messages by initially storing a received message in a queue and thereafter reading the message from the queue for processing. In addition, this second primary task performing means accumulates a count of the messages it reads from the queue. At the same time, or immediately thereafter, the primary task performing means sends the same messages to the second secondary task performing means which stores these messages in a message queue associated therewith. The messages stored in the queue of the second secondary task performing means are only processed if there is a failure of the second primary task performing means.
Other objects, the features and advantages of the invention will be apparent from the following detailed description when read in connection with the accompanying drawings which show, by way of example and not limitation, the apparatus for performing the invention.
FIG. 1 is a block diagram of a parallel computer system utilizing the invention; and
FIG. 2 is a generalized block diagram of one of the task processes in FIG. 1.
In FIG. 1 the parallel computer system PCS is shown including the primary task processor 11 with its associated backup or secondary task processor 12 and the second primary task processor 21 with its associated backup task processor 22 [A task processor can be a central processing unit (CPU), a program stored in a read only memory (ROM), and the state information stored in a random access memory (RAM) for performing a given process or task at a given time. The required hardware elements can be from the Motorola 68000 series]. These task processors are interconnected by means of the message bus MB which is preferably a dual message bus having complete redundancy. Furthermore, all task processors and particularly for the present application, the task processor 21 and its backup task processor 22 are shown connected via the message bus MB to a common memory CM (a common memory can be a pair of disc controllers connected to the pair of channels of the bus and rotating disc with sets of addressed registers on the disc being accessible to both processors of a pair). Although only two processor pairs, with one of the pairs being interconnected to a common memory are shown, it should be realized that innumerable processor pairs can be connected to the message bus MB. In an embodiment of the invention under construction, there are thirty-two such task processor pairs in the system. In the PCS messages can be transmitted from any task processor to any other task processor on the bus. In fact, the messages are transmitted from one task processor to other task processors in parallel over the two separate channels of the bus. In accordance with the invention, the messages are in general sent to three task processors, i.e., the originating task processor sends the message to the desired destination task processor as well as the backup task processor of the desired destination processor. Furthermore, this message is also preferably sent to the backup task processor of the originating task processor.
In FIG. 2, there is shown a generalized task processor XY. For the present description of the invention, the task processor XY can be considered to include a read counter RC, a write counter WC, a message queue MQ, and the remainder of the processor RP.
If the task processor is a primary processor, then the read counter RC is used and the write counter WC is ignored. If the task processor is a backup processor, then the write counter is used and the read counter is not used.
The read counter RC is used to accumulate a count of the number of messages read from the message queue MQ into the remainder of the processor RP. At the start of operations, this register is initialized and thereafter unit-incremented for each message read from the queue to the remainder of the processor. Whenever there is a synchronization between the primary processor and its associated backup processor, the read counter is initialized. As will hereinafter become apparent, synchronization is similar to checkpointing in that the primary processor and the backup processor are brought to the same state and, most particularly, that data spaces are identical.
The message that is transmitted by the originating processor is used in different ways by each of the receiving processors. The destination processor accepts the message and loads it into a queue for reading by the remainder of the processor. The message is also loaded into the queue of the backup processor of the receiving destination processor but is only for use by the remainder of the backup processor if there is a failure in the destination processor. Finally, the message is used in the backup processor of the sending processor merely to accumulate a count of the messages sent by the sending processor. The message is thereafter discarded (note that it is also possible for the backup processor of the destination processor to store a count of the messages transmitted by the originating processor for later use as will become apparent below).
There will now be described the fundamental operations of the system concerning the invention. Several assumptions will be made: first of all, it will be assumed that task processors 11 and 21 are primary task processors and task processors 12 and 22 are the respective backup task processors. Accordingly, task processors 11 and 21 do not need or utilize the write counter WC. Similarly, task processors 12 and 22 do not need nor utilize the read counter RC.
Before describing the fundamental operations, it will be useful to review the process of transmitting messages. A message in general includes a header, a body and an end-of-message indicator. Within the header there is at least the destination address of the message and the sending address of the message (it is preferably to incorporate in the header all three destination addresses, the addresses being based on position within the header). The conventional, well-known message interchange is for the sending unit to send a request to send the message to the destination unit. If the destination unit can receive the message it so tells the sending unit which then transmits the message. When the destination unit receives the message, it returns an acknowledgement message to the sending unit. Such a routine is well known in the art. For the present invention, it is only necessary to know that the sending processor "calls"(sends a request to send message) to all three processors. Only when all three processors (or in the case when the sending processor's backup is not used to count transmitted messages then only the two other processors are "listening", that is to say, indicate they can receive the message) will the sending processor transmit the message. Only when all processors have received the message and so acknowledged is the message transmission considered complete.
The following Table I shows the state of the pertinent element of the system prior to operation. As can be seen, the write counters WC of the task processors 12 and 22 (the backup processors) are initialized to a zero value. The read counters RC in the task processors 11 and 21 (the primary processors) are also initialized to zero. In addition, the message queues of all processors are empty.
TABLE I______________________________________TASK WRITE READ MESSAGEPROCESSOR COUNTER COUNTER QUEUE______________________________________11 unused 0 012 0 unused 021 unused 0 022 0 unused 0______________________________________
Assume that the task processor 11, a sending or originating primary task processor emits three messages onto the message bus MB. The messages are primarily destined for the receiving or destination primary task processor 21. As will be recalled above, these messages also are sent to the task processor 22, the secondary task processor for the destination task processor 21, and the task processor 12 (the secondary task processor for the sending task processor 11). After the three messages are sent, Table II shows the status of the relevant elements of the system. In particular, the write counter WC of the backup task processor 12 for the sending task processor 11 stores a count of three indicating that the sending task processor has sent three messages. The message queues of the destination task processors (primary and secondary) 21 and 22 each store the three messages as represented by the symbols M1, M2 and M3. Task processor 21 will use these messages for processing. Task processor 22 will hold these messages and only use them if there is a breakdown in task processor 21. Task processor 12 merely stores a count indicating that three messages have been transmitted by its primary task processor 11.
TABLE II______________________________________TASK WRITE READ MESSAGEPROCESSOR COUNTER COUNTER QUEUE______________________________________11 unused 0 012 3 unused 021 unused 0 M1, M2, M322 0 unused M1, M2, M3______________________________________
The next step in the operation is that the destination task processor 21 reads the messages M1 and M2 from its message queue and processes these messages. Table III shows the results of this operation. It should be noted that the difference now is that the message queue for task processor 21 shows merely the storage of the message M3 while its read counter stores a count of two indicating that two messages (M1 and M2) have been read from the message queue.
TABLE III______________________________________TASK WRITE READ MESSAGEPROCESSOR COUNTER COUNTER QUEUE______________________________________11 unused 0 012 3 unused 021 unused 2 M322 0 unused M1, M2, M3______________________________________
Assume in response thereto that the task processor 21 generates two messages M4 and M5 which it transmits to the task processor 11, now acting as a destination primary task processor. As can be seen from the following Table IV, the message queues for the task processors 11 and 12 each store the messages M4 and M5. In addition, the write counter of the task processor 22 accumulates a count of two.
TABLE IV______________________________________TASK WRITE READ MESSAGEPROCESSOR COUNTER COUNTER QUEUE______________________________________11 unused 0 M4, M512 3 unused M4, M521 unused 2 M322 2 unused M1, M2, M3______________________________________
During normal operation, a primary processor and its backup processor are synchronized whenever the queue of the saved messages in the backup processor becomes too large. At that time, the backup processor sends a special message to the primary processor requesting synchronization. In response thereto, the primary processor writes all the pages of its data space into a common memory shared with the backup processor. In addition, the primary processor sends the counter number of the number of messages it has read from its queue, either from the start of operation or thereafter from the previous synchronization, to the backup processor. The backup processor uses this count value to discard the corresponding number of messages from the start of its queue. In this way, should the backup processor start operating because of failure of the primary processor, it will only process messages that had not been processed since the previous synchronization. After sending the read count to the backup processor, the primary processor initializes its stored read count. In addition, the write counter in the backup processor is zeroed. It will be assumed that synchronization is called for and the following Table V shows the state of the pertinent elements after such synchronization.
TABLE V______________________________________TASK WRITE READ MESSAGEPROCESSOR COUNTER COUNTER QUEUE______________________________________11 unused 0 M4, M512 3 unused M4, M521 unused 0 M322 0 unused M3______________________________________
Note that the memory queues of the task processors 21 and 22 are identical and that the read counter of the task processor 21 shows a zero. What is not shown in the table is that the common memory CM2 stores the up-to-date data space of the task processor 21 which is now accessible to the task processor 22.
There will now be described the takeover of the processing by the backup task processor 22 because of the failure of the primary task processor 21. What should be realized is that the processor 22 will start processing at the point where the last synchronizing took place, or if the operation was just starting without any synchronizations, from the start of operations. It should be realized that either case is the same. To simplify the discussion, it will be assumed that there has been no synchronization and that the system is in the state represented by Table IV. At a failure, the backup processor takes over. It is preferable at the time of taking over that it spawn a new backup task processor; however, since this spawning does not form part of the present invention, that operation will not be discussed. However, the backup task processor starts processing the data. The important point is that in the processing, the backup processor 22, now acting as the primary processor, does not retransmit any messages that the now-failed primary processor 21 had emitted before failure. This is accomplished by means of of the count in the write counter of the processor 22. It will be noted that this, for the example given, contains a count of two indicating that the processor 21 before failing and from the time of starting or from the time of the last synchronization had emitted two messages. Therefore, the task processor 22 in the course of processing, when it reaches the first time to emit a message will not emit that message but decrement the write counter to a count of one. Thereafter, when the next message is to be emitted, again the message will be suppressed and the count decremented to zero. At that point, all messages previously emitted by processor 21 have been suppressed and all messages now generated by the processor 22 will be transmitted. Table VI shows the state at the time of failure.
TABLE VI______________________________________TASK WRITE READ MESSAGEPROCESSOR COUNTER COUNTER QUEUE______________________________________11 unused 0 M4, M512 3 unused M4, M521 unused failed failed22 2 unused M1, M2, M3______________________________________
There has thus been described the fundamental operations associated with the fault tolerant processing of messages. Although the fundamental operations have been given above in a particular order, it should be realized that they could occur in any order and could be among the various processors not specifically those cited in the example.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3735360 *||Aug 25, 1971||May 22, 1973||Ibm||High speed buffer operation in a multi-processing system|
|US4133027 *||Sep 13, 1977||Jan 2, 1979||Honeywell Inc.||Process control system with backup process controller|
|US4228496 *||Sep 7, 1976||Oct 14, 1980||Tandem Computers Incorporated||Multiprocessor system|
|US4333144 *||Feb 5, 1980||Jun 1, 1982||The Bendix Corporation||Task communicator for multiple computer system|
|US4347563 *||Jun 16, 1980||Aug 31, 1982||Forney Engineering Company||Industrial control system|
|US4351023 *||Apr 11, 1980||Sep 21, 1982||The Foxboro Company||Process control system with improved system security features|
|US4356546 *||Feb 5, 1980||Oct 26, 1982||The Bendix Corporation||Fault-tolerant multi-computer system|
|GB2032149A *||Title not available|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US4665520 *||Feb 1, 1985||May 12, 1987||International Business Machines Corporation||Optimistic recovery in a distributed processing system|
|US4709325 *||Sep 4, 1984||Nov 24, 1987||Nec Corporation||Loosely coupled multiprocessor system capable of transferring a control signal set by the use of a common memory|
|US4821170 *||Apr 17, 1987||Apr 11, 1989||Tandem Computers Incorporated||Input/output system for multiprocessors|
|US4823256 *||Jun 22, 1984||Apr 18, 1989||American Telephone And Telegraph Company, At&T Bell Laboratories||Reconfigurable dual processor system|
|US4873631 *||Apr 25, 1988||Oct 10, 1989||Ncr Corporation||Point of sale automatic back-up system and method|
|US4882669 *||Dec 9, 1987||Nov 21, 1989||Canon Kabushiki Kaisha||Multi computer fail safe control apparatus|
|US4954941 *||Aug 31, 1988||Sep 4, 1990||Bell Communications Research, Inc.||Method and apparatus for program updating|
|US4979108 *||Dec 26, 1989||Dec 18, 1990||Ag Communication Systems Corporation||Task synchronization arrangement and method for remote duplex processors|
|US4991174 *||Mar 28, 1989||Feb 5, 1991||Hitachi, Ltd.||Fault diagnostic distributed processing method and system|
|US5008817 *||Jul 22, 1988||Apr 16, 1991||Hitachi, Ltd.||Method and apparatus for transferring addresses and information in a buffer memory and a common main storage device|
|US5023778 *||Mar 23, 1990||Jun 11, 1991||General Motors Corporation||Interprocessor communication method|
|US5060138 *||Aug 31, 1990||Oct 22, 1991||Advanced Micro Devices, Inc.||Apparatus for use with a computing device for generating a substitute acknowledgement to an input when the computing device is in an operational hiatus|
|US5079740 *||Sep 21, 1989||Jan 7, 1992||Ncr Corporation||System and method of providing an automatic back-up primary terminal for a cluster of secondary terminals|
|US5086429 *||Apr 10, 1990||Feb 4, 1992||Honeywell Inc.||Fault-tolerant digital computing system with reduced memory redundancy|
|US5109330 *||Mar 24, 1987||Apr 28, 1992||Siemens Aktiengesellschaft||Multiprocessor system with sequential prioritized common memory access preventing memory update access by all processors except main processor|
|US5121488 *||Aug 14, 1990||Jun 9, 1992||International Business Machines Corporation||Sequence controller of an instruction processing unit for placing said unit in a ready, go, hold, or cancel state|
|US5122717 *||Oct 24, 1990||Jun 16, 1992||Toshiba Kikai Kabushiki Kaisha||Apparatus for producing NC programs using menu icons to define operations and producing method thereof|
|US5175828 *||Feb 13, 1989||Dec 29, 1992||Hewlett-Packard Company||Method and apparatus for dynamically linking subprogram to main program using tabled procedure name comparison|
|US5201040 *||Jun 20, 1988||Apr 6, 1993||Hitachi, Ltd.||Multiprocessor system having subsystems which are loosely coupled through a random access storage and which each include a tightly coupled multiprocessor|
|US5202980 *||Jul 2, 1990||Apr 13, 1993||Nec Corporation||Information processing system capable of readily taking over processing of a faulty processor|
|US5271013 *||May 9, 1990||Dec 14, 1993||Unisys Corporation||Fault tolerant computer system|
|US5276823 *||Mar 5, 1991||Jan 4, 1994||Tandem Computers Incorporated||Fault-tolerant computer system with redesignation of peripheral processor|
|US5285203 *||Jun 5, 1991||Feb 8, 1994||Matsuhita Electric Industrial Co., Ltd.||Message transmission device, and a message transmission system and a message transmission status advising system using the message transmission device|
|US5317726 *||Jun 26, 1991||May 31, 1994||Tandem Computers Incorporated||Multiple-processor computer system with asynchronous execution of identical code streams|
|US5339418 *||Dec 16, 1993||Aug 16, 1994||Digital Equipment Corporation||Message passing method|
|US5357612 *||Jan 11, 1993||Oct 18, 1994||International Business Machines Corporation||Mechanism for passing messages between several processors coupled through a shared intelligent memory|
|US5363503 *||Jan 22, 1992||Nov 8, 1994||Unisys Corporation||Fault tolerant computer system with provision for handling external events|
|US5388242 *||Nov 24, 1992||Feb 7, 1995||Tandem Computers Incorporated||Multiprocessor system with each processor executing the same instruction sequence and hierarchical memory providing on demand page swapping|
|US5459454 *||Sep 14, 1993||Oct 17, 1995||Matsushita Electric Industrial Co., Ltd.||Message transmission device, and a message transmission system and a message transmission status advising system using the message transmission device|
|US5459864 *||Feb 2, 1993||Oct 17, 1995||International Business Machines Corporation||Load balancing, error recovery, and reconfiguration control in a data movement subsystem with cooperating plural queue processors|
|US5463733 *||Aug 8, 1994||Oct 31, 1995||International Business Machines Corporation||Failure recovery apparatus and method for distributed processing shared resource control|
|US5488723 *||May 24, 1993||Jan 30, 1996||Cegelec||Software system having replicated objects and using dynamic messaging, in particular for a monitoring/control installation of redundant architecture|
|US5497497 *||Apr 22, 1993||Mar 5, 1996||Compaq Computer Corp.||Method and apparatus for resetting multiple processors using a common ROM|
|US5586249 *||Jul 12, 1995||Dec 17, 1996||Fujitsu Limited||Control information backup system|
|US5649152 *||Oct 13, 1994||Jul 15, 1997||Vinca Corporation||Method and system for providing a static snapshot of data stored on a mass storage system|
|US5675723 *||May 19, 1995||Oct 7, 1997||Compaq Computer Corporation||Multi-server fault tolerance using in-band signalling|
|US5687308 *||Jun 7, 1995||Nov 11, 1997||Tandem Computers Incorporated||Method to improve tolerance of non-homogeneous power outages|
|US5696895 *||Jun 19, 1995||Dec 9, 1997||Compaq Computer Corporation||Fault tolerant multiple network servers|
|US5729675 *||Aug 20, 1996||Mar 17, 1998||Compaq Computer Corporation||Apparatus for initializing a multiple processor computer system using a common ROM|
|US5764903 *||Sep 26, 1994||Jun 9, 1998||Acer America Corporation||High availability network disk mirroring system|
|US5781716 *||Feb 19, 1997||Jul 14, 1998||Compaq Computer Corporation||Fault tolerant multiple network servers|
|US5781741 *||Jun 29, 1995||Jul 14, 1998||Fujitsu Limited||Message communications system in a parallel computer|
|US5802265 *||Dec 1, 1995||Sep 1, 1998||Stratus Computer, Inc.||Transparent fault tolerant computer system|
|US5822512 *||Jun 5, 1996||Oct 13, 1998||Compaq Computer Corporartion||Switching control in a fault tolerant system|
|US5828569 *||May 21, 1997||Oct 27, 1998||Tandem Computers, Inc.||Method and apparatus for maintaining network connections across a voluntary process switchover|
|US5835953 *||Nov 8, 1996||Nov 10, 1998||Vinca Corporation||Backup system that takes a snapshot of the locations in a mass storage device that has been identified for updating prior to updating|
|US5864657 *||Nov 29, 1995||Jan 26, 1999||Texas Micro, Inc.||Main memory system and checkpointing protocol for fault-tolerant computer system|
|US5867703 *||Jan 27, 1998||Feb 2, 1999||Compaq Computer Corporation||Common reset ROM|
|US5901281 *||May 3, 1995||May 4, 1999||Hitachi, Ltd.||Processing unit for a computer and a computer system incorporating such a processing unit|
|US5928368 *||Jun 23, 1994||Jul 27, 1999||Tandem Computers Incorporated||Method and apparatus for fault-tolerant multiprocessing system recovery from power failure or drop-outs|
|US5991518 *||Jan 28, 1997||Nov 23, 1999||Tandem Computers Incorporated||Method and apparatus for split-brain avoidance in a multi-processor system|
|US6012150 *||Mar 27, 1997||Jan 4, 2000||International Business Machines Corporation||Apparatus for synchronizing operator initiated commands with a failover process in a distributed processing system|
|US6023772 *||Jan 24, 1997||Feb 8, 2000||Hewlett-Packard Company||Fault-tolerant processing method|
|US6032271 *||Jun 5, 1996||Feb 29, 2000||Compaq Computer Corporation||Method and apparatus for identifying faulty devices in a computer system|
|US6035415 *||Jan 24, 1997||Mar 7, 2000||Hewlett-Packard Company||Fault-tolerant processing method|
|US6101599 *||Jun 29, 1998||Aug 8, 2000||Cisco Technology, Inc.||System for context switching between processing elements in a pipeline of processing elements|
|US6119215 *||Jun 29, 1998||Sep 12, 2000||Cisco Technology, Inc.||Synchronization and control system for an arrayed processing engine|
|US6173386||Dec 14, 1998||Jan 9, 2001||Cisco Technology, Inc.||Parallel processor with debug capability|
|US6195739||Jun 29, 1998||Feb 27, 2001||Cisco Technology, Inc.||Method and apparatus for passing data among processor complex stages of a pipelined processing engine|
|US6216236||Nov 10, 1998||Apr 10, 2001||Tokyo, Japan||Processing unit for a computer and a computer system incorporating such a processing unit|
|US6223304 *||Jun 18, 1998||Apr 24, 2001||Telefonaktiebolaget Lm Ericsson (Publ)||Synchronization of processors in a fault tolerant multi-processor system|
|US6237108 *||Oct 2, 1997||May 22, 2001||Fujitsu Limited||Multiprocessor system having redundant shared memory configuration|
|US6260155||May 1, 1998||Jul 10, 2001||Quad Research||Network information server|
|US6272621||Aug 18, 2000||Aug 7, 2001||Cisco Technology, Inc.||Synchronization and control system for an arrayed processing engine|
|US6314515||Jul 19, 1999||Nov 6, 2001||Compaq Computer Corporation||Resetting multiple processors in a computer system|
|US6385747||Dec 14, 1998||May 7, 2002||Cisco Technology, Inc.||Testing of replicated components of electronic device|
|US6438707 *||Aug 10, 1999||Aug 20, 2002||Telefonaktiebolaget Lm Ericsson (Publ)||Fault tolerant computer system|
|US6442669||Nov 30, 2000||Aug 27, 2002||Cisco Technology, Inc.||Architecture for a process complex of an arrayed pipelined processing engine|
|US6463529||Feb 10, 1997||Oct 8, 2002||Compaq Computer Corporation, Inc.||Processor based system with system wide reset and partial system reset capabilities|
|US6487464 *||Feb 25, 2000||Nov 26, 2002||Intel Corporation||Power-on software for robust boot|
|US6496940 *||Jun 7, 1995||Dec 17, 2002||Compaq Computer Corporation||Multiple processor system with standby sparing|
|US6505269||May 16, 2000||Jan 7, 2003||Cisco Technology, Inc.||Dynamic addressing mapping to eliminate memory resource contention in a symmetric multiprocessor system|
|US6513108||Jun 29, 1998||Jan 28, 2003||Cisco Technology, Inc.||Programmable processing engine for efficiently processing transient data|
|US6529983||Nov 3, 1999||Mar 4, 2003||Cisco Technology, Inc.||Group and virtual locking mechanism for inter processor synchronization|
|US6633996||Apr 13, 2000||Oct 14, 2003||Stratus Technologies Bermuda Ltd.||Fault-tolerant maintenance bus architecture|
|US6662252||Dec 8, 2002||Dec 9, 2003||Cisco Technology, Inc.||Group and virtual locking mechanism for inter processor synchronization|
|US6681341||Nov 3, 1999||Jan 20, 2004||Cisco Technology, Inc.||Processor isolation method for integrated multi-processor systems|
|US6687851||Apr 13, 2000||Feb 3, 2004||Stratus Technologies Bermuda Ltd.||Method and system for upgrading fault-tolerant systems|
|US6691257||Apr 13, 2000||Feb 10, 2004||Stratus Technologies Bermuda Ltd.||Fault-tolerant maintenance bus protocol and method for using the same|
|US6708283||Apr 13, 2000||Mar 16, 2004||Stratus Technologies, Bermuda Ltd.||System and method for operating a system with redundant peripheral bus controllers|
|US6728839||Oct 28, 1998||Apr 27, 2004||Cisco Technology, Inc.||Attribute based memory pre-fetching technique|
|US6735715||Apr 13, 2000||May 11, 2004||Stratus Technologies Bermuda Ltd.||System and method for operating a SCSI bus with redundant SCSI adaptors|
|US6766413||Mar 1, 2001||Jul 20, 2004||Stratus Technologies Bermuda Ltd.||Systems and methods for caching with file-level granularity|
|US6766479||Feb 28, 2001||Jul 20, 2004||Stratus Technologies Bermuda, Ltd.||Apparatus and methods for identifying bus protocol violations|
|US6769027 *||Jan 31, 2000||Jul 27, 2004||Avaya Technology Corp.||System and method for using multi-headed queues for bookmarking in backup/recover scenarios|
|US6799224||Mar 10, 1999||Sep 28, 2004||Quad Research||High speed fault tolerant mass storage network information server|
|US6802022||Sep 18, 2000||Oct 5, 2004||Stratus Technologies Bermuda Ltd.||Maintenance of consistent, redundant mass storage images|
|US6820213||Apr 13, 2000||Nov 16, 2004||Stratus Technologies Bermuda, Ltd.||Fault-tolerant computer system with voter delay buffer|
|US6836838||Aug 16, 2002||Dec 28, 2004||Cisco Technology, Inc.||Architecture for a processor complex of an arrayed pipelined processing engine|
|US6862689||Apr 12, 2001||Mar 1, 2005||Stratus Technologies Bermuda Ltd.||Method and apparatus for managing session information|
|US6874102||Mar 5, 2001||Mar 29, 2005||Stratus Technologies Bermuda Ltd.||Coordinated recalibration of high bandwidth memories in a multiprocessor computer|
|US6892237||Mar 28, 2000||May 10, 2005||Cisco Technology, Inc.||Method and apparatus for high-speed parsing of network messages|
|US6920562||Dec 18, 1998||Jul 19, 2005||Cisco Technology, Inc.||Tightly coupled software protocol decode with hardware data encryption|
|US6948010||Dec 20, 2000||Sep 20, 2005||Stratus Technologies Bermuda Ltd.||Method and apparatus for efficiently moving portions of a memory block|
|US6954877||Nov 29, 2001||Oct 11, 2005||Agami Systems, Inc.||Fault tolerance using logical checkpointing in computing systems|
|US6996750||May 31, 2001||Feb 7, 2006||Stratus Technologies Bermuda Ltd.||Methods and apparatus for computer bus error termination|
|US7028218 *||Dec 2, 2002||Apr 11, 2006||Emc Corporation||Redundant multi-processor and logical processor configuration for a file server|
|US7065672||Mar 28, 2001||Jun 20, 2006||Stratus Technologies Bermuda Ltd.||Apparatus and methods for fault-tolerant computing using a switching fabric|
|US7168001 *||Feb 6, 2004||Jan 23, 2007||Hewlett-Packard Development Company, L.P.||Transaction processing apparatus and method|
|US7185224||Dec 10, 2003||Feb 27, 2007||Cisco Technology, Inc.||Processor isolation technique for integrated multi-processor systems|
|US7228459 *||May 19, 2003||Jun 5, 2007||Tellabs Petaluma, Inc.||Apparatus and method that provides a primary server and a backup server that both support a RADIUS client and share an IP address|
|US7263476||Jun 12, 2000||Aug 28, 2007||Quad Research||High speed information processing and mass storage system and method, particularly for information and application servers|
|US7368835 *||Apr 21, 2005||May 6, 2008||Tellabs Bedford, Inc.||Power distribution system and method|
|US7370248 *||Nov 7, 2003||May 6, 2008||Hewlett-Packard Development Company, L.P.||In-service raid mirror reconfiguring|
|US7380101||Dec 27, 2004||May 27, 2008||Cisco Technology, Inc.||Architecture for a processor complex of an arrayed pipelined processing engine|
|US7386851 *||Jan 4, 2008||Jun 10, 2008||International Business Machines Corporation||System and method for implementing dynamic lifetime reliability extension for microprocessor architectures|
|US7395332||Mar 21, 2005||Jul 1, 2008||Cisco Technology, Inc.||Method and apparatus for high-speed parsing of network messages|
|US7472038||Apr 16, 2007||Dec 30, 2008||International Business Machines Corporation||Method of predicting microprocessor lifetime reliability using architecture-level structure-aware techniques|
|US7472313 *||Mar 13, 2007||Dec 30, 2008||Tellabs Petaluma, Inc.||Primary server and backup server that share an IP address and a limited number of message identifiers|
|US7493512 *||Oct 4, 2005||Feb 17, 2009||First Data Corporation||System and method for providing data services via a network|
|US7895412||Jun 27, 2002||Feb 22, 2011||Cisco Tehnology, Inc.||Programmable arrayed processing engine architecture for a network switch|
|US8140887||Feb 11, 2009||Mar 20, 2012||First Data Corporation||System and method for providing data services via a network|
|US8909816 *||Mar 19, 2012||Dec 9, 2014||Kaminario Technologies Ltd.||Implementing a logical unit reset command in a distributed storage system|
|US20020116555 *||Dec 20, 2000||Aug 22, 2002||Jeffrey Somers||Method and apparatus for efficiently moving portions of a memory block|
|US20020124202 *||Mar 5, 2001||Sep 5, 2002||John Doody||Coordinated Recalibration of high bandwidth memories in a multiprocessor computer|
|US20020144175 *||Mar 28, 2001||Oct 3, 2002||Long Finbarr Denis||Apparatus and methods for fault-tolerant computing using a switching fabric|
|US20020194548 *||May 31, 2001||Dec 19, 2002||Mark Tetreault||Methods and apparatus for computer bus error termination|
|US20030028584 *||Jul 15, 2002||Feb 6, 2003||Mark Coniglio||System and method for providing network management|
|US20040078652 *||Mar 8, 2002||Apr 22, 2004||Tapper Gunnar D.||Using process quads to enable continuous services in a cluster environment|
|US20040153749 *||Dec 2, 2002||Aug 5, 2004||Schwarm Stephen C.||Redundant multi-processor and logical processor configuration for a file server|
|US20040199812 *||Nov 29, 2001||Oct 7, 2004||Earl William J.||Fault tolerance using logical checkpointing in computing systems|
|US20040250173 *||May 19, 2003||Dec 9, 2004||Jiang Tsang Ming||Apparatus and method that provides a primary server and a backup server that both support a RADIUS client and share an IP address|
|US20050033916 *||Sep 20, 2004||Feb 10, 2005||Richard Dellacona||High speed fault tolerant mass storage network information server|
|US20050102603 *||Nov 7, 2003||May 12, 2005||Gunnar Tapper||In-service raid mirror reconfiguring|
|US20050132250 *||Dec 16, 2003||Jun 16, 2005||Hewlett-Packard Development Company, L.P.||Persistent memory device for backup process checkpoint states|
|US20050165966 *||Mar 21, 2005||Jul 28, 2005||Silvano Gai||Method and apparatus for high-speed parsing of network messages|
|US20050187891 *||Feb 6, 2004||Aug 25, 2005||Johnson Charles S.||Transaction processing apparatus and method|
|US20050193059 *||Apr 28, 2005||Sep 1, 2005||Richard Dellacona||High speed fault tolerant mass storage network information server|
|US20050203989 *||May 9, 2005||Sep 15, 2005||Richard Dellacona||High speed information processing and mass storage system and method, particularly for information and application servers|
|US20050216552 *||Mar 24, 2004||Sep 29, 2005||Samuel Fineberg||Communication-link-attached persistent memory system|
|US20050223146 *||Feb 4, 2005||Oct 6, 2005||Richard Dellacona||High speed information processing and mass storage system and method, particularly for information and application servers|
|US20050251295 *||Apr 21, 2005||Nov 10, 2005||Tellabs Bedford||Power distribution system and method|
|US20060222125 *||Mar 31, 2005||Oct 5, 2006||Edwards John W Jr||Systems and methods for maintaining synchronicity during signal transmission|
|US20060222126 *||Jun 2, 2005||Oct 5, 2006||Stratus Technologies Bermuda Ltd.||Systems and methods for maintaining synchronicity during signal transmission|
|US20070028144 *||Jul 29, 2005||Feb 1, 2007||Stratus Technologies Bermuda Ltd.||Systems and methods for checkpointing|
|US20070075747 *||Oct 4, 2005||Apr 5, 2007||Jacob Apelbaum||System and method for providing data services via a network|
|US20070076228 *||Oct 4, 2005||Apr 5, 2007||Jacob Apelbaum||System and method for providing data services via a network|
|US20070174729 *||Mar 13, 2007||Jul 26, 2007||Jiang Tsang M||Primary server and backup server that share an IP address and a limited number of message identifiers|
|US20070266074 *||Jul 19, 2007||Nov 15, 2007||Quad Research|
|US20080172573 *||Dec 19, 2007||Jul 17, 2008||Saab Ab||Method for ensuring backup function to an electrical system in a vehicle and an electrical system as such|
|US20090013207 *||Aug 11, 2008||Jan 8, 2009||International Business Machines Corporation||Predicting microprocessor lifetime reliability using architecture-level structure-aware techniques|
|US20090178051 *||May 9, 2008||Jul 9, 2009||International Business Machines Corporation||Method for implementing dynamic lifetime reliability extension for microprocessor architectures|
|US20130024015 *||Feb 1, 2012||Jan 24, 2013||Byung Hoon Yoo||Method for controlling building automatic controll device|
|US20130246660 *||Mar 19, 2012||Sep 19, 2013||Kaminario Technologies Ltd.||Implementing a logical unit reset command in a distributed storage system|
|EP1936502A1||Dec 19, 2006||Jun 25, 2008||Saab Ab||A method for ensuring backup function to an electrical system in a vehicle and an electrical system as such|
|EP2846265A1||Aug 23, 2010||Mar 11, 2015||Chicago Mercantile Exchange, Inc.||Match server for a financial exchange having fault tolerant operation|
|WO2003010869A1 *||Jul 15, 2002||Feb 6, 2003||Shape Of Time Inc||System and method for providing network management|
|WO2007044246A2 *||Sep 29, 2006||Apr 19, 2007||Jacob Apelbaum||System and method for providing data services via a network|
|U.S. Classification||714/13, 714/E11.08, 714/E11.071, 714/E11.117|
|International Classification||G06F15/16, G06F15/177, G06F11/20, G06F11/14|
|Cooperative Classification||G06F11/2038, G06F11/2005, G06F11/2007, G06F11/2043, G06F11/2097|
|European Classification||G06F11/14A, G06F11/20, G06F11/20C4, G06F11/20C2, G06F11/20U|
|Nov 23, 1982||AS||Assignment|
Owner name: PARALLEL COMPUTER SYSTEMS, INC., 210 SYLVAN AVE.,
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNORS:GLAZER, SAM D.;BAUMBACH, JAMES;BORG, ANITA;AND OTHERS;REEL/FRAME:004071/0458
Effective date: 19821122
|Feb 15, 1990||REMI||Maintenance fee reminder mailed|
|May 20, 1990||LAPS||Lapse for failure to pay maintenance fees|
|Jul 31, 1990||FP||Expired due to failure to pay maintenance fee|
Effective date: 19900520